README.rst
.. image:: https://img.shields.io/travis/Nekmo/simple-monitor-alert.svg?style=flat-square
:target: https://travis-ci.org/Nekmo/simple-monitor-alert
:alt: Latest Travis CI build status
.. image:: https://img.shields.io/pypi/v/simple-monitor-alert.svg?style=flat-square
:target: https://pypi.python.org/pypi/simple-monitor-alert
:alt: Latest PyPI version
.. image:: https://img.shields.io/pypi/pyversions/simple-monitor-alert.svg?style=flat-square
:target: https://pypi.python.org/pypi/simple-monitor-alert
:alt: Python versions
.. image:: https://img.shields.io/codeclimate/github/Nekmo/simple-monitor-alert.svg?style=flat-square
:target: https://codeclimate.com/github/Nekmo/simple-monitor-alert
:alt: Code Climate
.. image:: https://img.shields.io/codecov/c/github/Nekmo/simple-monitor-alert/master.svg?style=flat-square
:target: https://codecov.io/github/Nekmo/simple-monitor-alert
:alt: Test coverage
.. image:: https://img.shields.io/requires/github/Nekmo/simple-monitor-alert.svg?style=flat-square
:target: https://requires.io/github/Nekmo/simple-monitor-alert/requirements/?branch=master
:alt: Requirements Status
Simple Monitor Alert
####################
A simple monitor with alerts for Unix/Linux under the KISS philosophy. Keep It Simple, Stupid!
- **Light**: Only ~7MiB of RAM. (That's great for your raspberry pi!)
- Very **easy** to use and understand.
- Write **your own monitors** in any language (Bash, Python, Perl, JS, Ruby, PHP, C, C++...).
- **Awesome features**: send alerts once or several times, graphic peak...
- No server required. You can run as a **daemon or using crond**.
- Easy to **debug** and test.
- Multiple ways to send alerts: **email, telegram**...
- **Easy configuration** in a single file.
Available monitors:
- **Hard disks**: SMART, temperature, free space, Mdadm (linux RAID).
- **System**: CPU usage, free RAM, free SWAP, monitorize services (daemons), sensors, [time]...
- **Web**: load time, return code, content in page.
- **Network**: ping, [port, dynamic ip].
- **Others**: [Log monitor, google cloud print]
Screencast
==========
.. image:: https://asciinema.org/a/ez93g4bewogf6wss7bxxnc5tz.png
:target: https://asciinema.org/a/ez93g4bewogf6wss7bxxnc5tz
3-Step Quick Start Guide
========================
1. Install it from Pypi::
$ sudo pip install simple-monitor-alert
2. Edit `/etc/simple-monitor-alert/sma.ini` and defines the recipient of the alerts::
[mail]
to = awesome@email.com
3. Execute sma::
# Just once:
$ sma
# or... Run as a service (daemon)
$ sma service
# or... Run usign system service:
$ sudo systemctl start sma
And yes, that's it!
5 minutes guide
===============
Files and directories:
* `/etc/simple-monitor-alert/sma.ini` (file): all-in-one config file. Configure monitors and alerts methods.
* `/etc/simple-monitor-alert/monitors-available` (directory): All monitors available for usage. You can create monitors here.
* `/etc/simple-monitor-alert/monitors-enabled` (directory): All monitors that are here are activated.
* `/etc/simple-monitor-alert/alerts` (directory): Alerts methods available. You need to configure them in sma.ini.
Enable and disable monitors
---------------------------
All monitors in `/etc/simple-monitor-alert/monitors-enabled` are enabled. It is recommended that files are symbolic
links. To **activate** a monitor::
$ cd /etc/simple-monitor-alert/monitors-enabled
$ sudo ln -s ../monitors-available/mdadm.sh
To **disable**::
$ cd /etc/simple-monitor-alert/monitors-enabled
$ rm mdadm.sh # It's safe. mdadm is a symlink.
We recommend you read the beginning of the monitor before activating. Some monitors may require parameters and
configure the system. For example::
$ head -n 6 /etc/simple-monitor-alert/monitors-available/service.sh
#!/usr/bin/env bash
# Service Status monitor.
# Verify that the service is running.
# It requires a parameter: service name. For example, sshd.
# [service]
# service_status.param = sshd
To pass the parameter you must add the following to `sma.ini`::
[service]
service_status.param = sshd
To monitor multiple services::
[service]
service_status(sshd).param = sshd
service_status(ntpd).param = ntpd
Debugging
---------
You can test your monitors running them::
$ /etc/simple-monitor-alert/monitors-available/mdadm.sh
mdadm(md0).name = 'Mdadm /dev/md0'
mdadm(md0).expected = 0
mdadm(md0).value = 0
You can also run sma and see the results::
$ sma
2016-05-03 00:28:14,972 - sma - INFO - Trigger: [success] (mdadm) mdadm(md0). Result: 0 == 0
2016-05-03 00:28:14,990 - sma - INFO - Trigger: [success] (system) ram. Result: 32.1427 <= 85
2016-05-03 00:28:14,990 - sma - INFO - Trigger: [success] (system) cpu. Result: 9.57627 <= 80
2016-05-03 00:28:15,156 - sma - WARNING - Trigger: [warning] (hdds) pcnt_use(sdc1). Assertion 98 <= 80 failed.
Extra info: Space: 23G/25G
2016-05-03 00:28:15,157 - sma - WARNING - Trigger: [warning] (hdds) pcnt_use(md0). Assertion 100 <= 80 failed.
Extra info: Space: 5,4T/5,5T
To test the alerts you can use::
$ sma alerts --test
My first monitor
----------------
SMA works by checking the output of your monitor script. A monitor has observables. Each observable has 2 major
sections: the expected value and the value obtained::
observable1.expected = yes
observable1.value = yes
observable2.expected = yes
observable2.value = no
In this example the first observable is fine and the second is under error. Your program should return something
similar. The following example check that a file exists::
#!/usr/bin/env bash
echo "file_exists.expected = yes"
if [ -f "/path/to/file" ]; then value="yes"; else value="no"; fi
echo "file_exists.value = $value"
Output::
$ /etc/simple-monitor-alert/monitors-available/example-1.sh
file_exists.expected = yes
file_exists.value = no
There are more options with monitors, such as obtaining arguments. For more information see the documentation.